Howdy,
I'm pretty deep in the rabbit hole trying to figure out a slot port design for a vented car audio sub (JL12W7AE3). I've got a few questions and will take any single answer to any of them as a blessing.
So JL recommends slotted ports over round for whatever reason and I'm in the ballpark of a 2.25"x11.5" port area. It's a fairly typical box design with one wall of the box serving as one wall of the port and an extra wall inside the enclosure to box in the vent. The port length based on some calculators is in the range of 30" for fb=32Hz which means the port has to make one turn around a corner of the enclosure to get the Lv I need given the overall dimensions of the box (box is ~26"x13"x16" but the 26" is what I can adjust most easily). But then there's a ton of uncertainty in the Lv. So, questions:
1) I can't find a port length formula without baked-in units. One post on this forum actually uses c^2/(16pi) but has a factor of 10 in it that I suspect is a conversion factor (when I use it with units of inches, my port length comes out ~10x too long). Others assume SI units (2.35625e4 is 10c^2/(16pi) in m/s). I stared at my math a million times (more or less) and can't find an error, so I suspect that factor of 10, but you tell me. For now, I'll use the 2.35625e4 and convert the Lv result from m to in.
2) At least one online vent calculator distinguishes between slot, square and round vent shapes. It also gives different solutions with the same effective D, so the underlying equations are different. Of course, the equations aren't given. For square or slot, I have to put in the port area's width and length, so maybe aspect ratio is involved. Dunno. Any intuition on that? I'm only 5:1 with my port area dimensions and these are the same as the "recommended ported enclosure" port specs from JL (just being safe), though their box is too cubic for what I want. Many say just compute effective diameter from slot area and pretend it's round, but I'm not convinced. Assuming it's round gets less obvious when I consider k.
3) k=?. Since vents have two ends, I have to figure out my effective k by summing the k for both ends, and I have different questions about each end, so...
3a) k at the vent mouth: I see plenty of stuff about k for tube ports, great stuff all over. I'm going to add a 3/4" radius bezel/flare around the port mouth, so I'm guessing maybe I just add 3/8" to Lv if I even bother (aside from the k correction). The slot mouth is on the same box face as the sub baffle and the slot is vertical relative to the floor of the vehicle - on purpose; if I put the slot under the sub, it makes the box taller and then the floor affects the effective port length. So the vent opens and next to one long edge of the mouth is the speaker. There's "nothing" next to the other long edge of the mouth (no baffle, just the 3/4" radius flare on the outside wall of the box). If I had a baffle all the way around the mouth, I'd have one value for k. If I don't, I have another k. With the speaker on one side and no baffle on the other, I'm not sure. Does the speaker side count as a baffle? If it's playing and resonating near fb, I'd think the in-phase pressure from the sub would act as an extension of the inner vent wall to some extent. What do you do with a slot that has two different conditions on each of its long edges?
3b) Inside the box, on the other end of the port, the port makes a turn and then ends part-way along the other enclosure wall, the one opposite the sub baffle. On the inner side, I'll again have a 3/4" radius flare that wraps 270 degrees around the end of the interior wall, but the other wall extends many more inches. Again, I can add 3/8" to Lv for the flare (independent of k), but what k do I use? JL says add W/2 where W is 2.25" in my case. Obviously, this doesn't play nice with the k*D term in the Lv equation because it doesn't account for a rectangular port. The W/2 stuff makes some intuitive sense. I could pick any W I want, in theory, and get the same effective D, so there can't be one k value to rule them all. Aside from having one side with a wall and the other without, the interior wall also lacks a baffle; it just wraps around the 3/4" radius flare. So k is particularly a mystery here.
So I've got myself wrapped around the axle. I think slot ported requires some special treatment because it's so far from having a circular cross section and a single k*D doesn't make intuitive sense, particular with these end conditions. I can cop out and use the JL vented enclosure port length and W/2 guidance, maybe compensating a tad bit for the radiused flares, but I'm a scientist and I really want to understand this a lot better.
Thanks in advance!
-Mack
I'm pretty deep in the rabbit hole trying to figure out a slot port design for a vented car audio sub (JL12W7AE3). I've got a few questions and will take any single answer to any of them as a blessing.
So JL recommends slotted ports over round for whatever reason and I'm in the ballpark of a 2.25"x11.5" port area. It's a fairly typical box design with one wall of the box serving as one wall of the port and an extra wall inside the enclosure to box in the vent. The port length based on some calculators is in the range of 30" for fb=32Hz which means the port has to make one turn around a corner of the enclosure to get the Lv I need given the overall dimensions of the box (box is ~26"x13"x16" but the 26" is what I can adjust most easily). But then there's a ton of uncertainty in the Lv. So, questions:
1) I can't find a port length formula without baked-in units. One post on this forum actually uses c^2/(16pi) but has a factor of 10 in it that I suspect is a conversion factor (when I use it with units of inches, my port length comes out ~10x too long). Others assume SI units (2.35625e4 is 10c^2/(16pi) in m/s). I stared at my math a million times (more or less) and can't find an error, so I suspect that factor of 10, but you tell me. For now, I'll use the 2.35625e4 and convert the Lv result from m to in.
2) At least one online vent calculator distinguishes between slot, square and round vent shapes. It also gives different solutions with the same effective D, so the underlying equations are different. Of course, the equations aren't given. For square or slot, I have to put in the port area's width and length, so maybe aspect ratio is involved. Dunno. Any intuition on that? I'm only 5:1 with my port area dimensions and these are the same as the "recommended ported enclosure" port specs from JL (just being safe), though their box is too cubic for what I want. Many say just compute effective diameter from slot area and pretend it's round, but I'm not convinced. Assuming it's round gets less obvious when I consider k.
3) k=?. Since vents have two ends, I have to figure out my effective k by summing the k for both ends, and I have different questions about each end, so...
3a) k at the vent mouth: I see plenty of stuff about k for tube ports, great stuff all over. I'm going to add a 3/4" radius bezel/flare around the port mouth, so I'm guessing maybe I just add 3/8" to Lv if I even bother (aside from the k correction). The slot mouth is on the same box face as the sub baffle and the slot is vertical relative to the floor of the vehicle - on purpose; if I put the slot under the sub, it makes the box taller and then the floor affects the effective port length. So the vent opens and next to one long edge of the mouth is the speaker. There's "nothing" next to the other long edge of the mouth (no baffle, just the 3/4" radius flare on the outside wall of the box). If I had a baffle all the way around the mouth, I'd have one value for k. If I don't, I have another k. With the speaker on one side and no baffle on the other, I'm not sure. Does the speaker side count as a baffle? If it's playing and resonating near fb, I'd think the in-phase pressure from the sub would act as an extension of the inner vent wall to some extent. What do you do with a slot that has two different conditions on each of its long edges?
3b) Inside the box, on the other end of the port, the port makes a turn and then ends part-way along the other enclosure wall, the one opposite the sub baffle. On the inner side, I'll again have a 3/4" radius flare that wraps 270 degrees around the end of the interior wall, but the other wall extends many more inches. Again, I can add 3/8" to Lv for the flare (independent of k), but what k do I use? JL says add W/2 where W is 2.25" in my case. Obviously, this doesn't play nice with the k*D term in the Lv equation because it doesn't account for a rectangular port. The W/2 stuff makes some intuitive sense. I could pick any W I want, in theory, and get the same effective D, so there can't be one k value to rule them all. Aside from having one side with a wall and the other without, the interior wall also lacks a baffle; it just wraps around the 3/4" radius flare. So k is particularly a mystery here.
So I've got myself wrapped around the axle. I think slot ported requires some special treatment because it's so far from having a circular cross section and a single k*D doesn't make intuitive sense, particular with these end conditions. I can cop out and use the JL vented enclosure port length and W/2 guidance, maybe compensating a tad bit for the radiused flares, but I'm a scientist and I really want to understand this a lot better.
Thanks in advance!
-Mack
Last edited:
The effective length of a port is larger than its physical length. Factor 'k' is used to describe the correction. It depends on various things like nearby cabinet walls, whether the port has a round or square cross section and whether it has flares at the ends. How the calculator handles this correction is not always indicated explicitly.
Visaton Boxsim contains a nice calculator that includes the effect of nearby cabinet walls on the end correction.
Visaton Boxsim contains a nice calculator that includes the effect of nearby cabinet walls on the end correction.
I'm in heaven GM! Thanks a bunch! I'll be digesting this for a couple of hours. The notion of splitting the slot has given me a bunch of ideas that will make the woodworking side of it super fun! Got 100 square feet of Baltic birch 18mm plywood today to stock up on project materials for the holidays. Just need to nail down the design for my W7. Thanks a million!
TBTL: Thanks for the suggestion! I started playing with it but still need to look up a tutorial or three. I already got yelled at in German by the software a couple of times. Working in cm and liters is a bit annoying (need to upgrade my spreadsheet to work in general length units). Wish everyone used SI units, but, alas, I was raised with inches and pounds. Anyway, I'm building a fair arsenal of software!
-Mack
TBTL: Thanks for the suggestion! I started playing with it but still need to look up a tutorial or three. I already got yelled at in German by the software a couple of times. Working in cm and liters is a bit annoying (need to upgrade my spreadsheet to work in general length units). Wish everyone used SI units, but, alas, I was raised with inches and pounds. Anyway, I'm building a fair arsenal of software!
-Mack
Lots of awesome info in that, thanks again!
(1) The K values (look like something from Boxsim) for various port configurations (with ports supported by 0, 1, 2 or 3 enclosure walls) appear to me to be the combined correction for both ends. I'm guessing it is both ends based on the following logic: The write-up has 0.613 for free and 0.85 for baffle but these are radius-based, not diameter, so the typical k in the port k*D correction term would be half these values (.3065 for free and .425 for baffle). The sum of those is 0.7315 which would be the total k for baffle on one end and free on the other which jives with the Boxsim k=0.732 for a port away from all enclosure walls. Am I getting that right?
(2) If that's the case, any idea what the k values are for each end in the other 3 cases? The inner port end has a different boundary condition than the outer end. For example, a square port in the corner of the enclosure (total k=1.728) has for each of the four sides of the square: Inside - 2 walls and 2 free, Outside - 2 free and 2 baffle.
(3) If I divide the slot, do I use D for one port in the kD correction or the total effective D for the whole slot?
(4) Do you agree with the k~=1.5 in that write-up for the example where the slot port is divided into four ports? The write-up assumes that two of the ports (1st and 4th) are in the corners each with k=1.728 and two (2nd and 3rd) are along only one side each with k=1.23. The effective k is assumed to be the average of the four.
I would think the divided port would still act like a single slot port with k=2.25, but I'm obviously still a newbie. Some scenarios:
- A single slot at 2.25"x11.5" has equivalent hydraulic diameter D=5.16" and k=2.227.
- If I divide it into two with 1/2" material, each sub-port has D=3.75" and (maybe) an average k=1.728 (both sub-ports are in corners of the enclosure).
- Dividing it into 3 gives D=3" and k=1.562.
If I use the the above D values in kD, I get length corrections:
- 0 dividers: 11.5"
- 1 divider: 6.5"
- 2 dividers: 4.76"
This just by adding dividers?!
I'm sure someone has slid a divider into a single slot port and either got a different FR or didn't. I haven't and I don't know (obviously).
Thanks!
-Mack
You're welcome!
Indeed! Bjorno was quite the complete pro designer with the necessary math skills to work it out using the wave equation and the tools/personnel at Dynaudio? to go way beyond my mostly Altec 'enhanced' DIY efforts and somewhat of an advanced mentor for me, but sadly 'left the building' recently.
1] Correctomundo!
2, 3, 4] Not anytime soon, if ever. Ages ago I figured them all out based on Altec data [actually Western Electric's] and at this late date just remember they didn't agree WRT any previously slot/shelf port design theory, but it's all on two damaged HDs, so probably lost to me and without the data can't compare to this one or try again as it was absorbed by EV and they don't share, at least to me.
I did notice though that he changed/added a significant amount of info from earlier versions of his and others and had planned to discuss it with him after the holidays, especially #4 since Altec lumped them together as one, though used thin mica dividers to mimic WE's metal ones with a high aspect ratio end correction [added friction, re heat/AC duct design], though wondering if NACA duct design might be a better choice, so with typical ~3/4" thick defining square to modest rectangular [1:1.273 Altec max] 1/2 WL open pipes it seems reasonable that it might be more like his .......... regardless, based on who made it, I'm inclined to accept this latest one as at least 'close enough' till proven otherwise.
With mica dividers, no measurable difference Vs no dividers with the same tuning or round vent of the same area, though need to keep in mind that my measurement capability was only a notch above none compared to the Altec distributor's $20+ k [early '70s dollars] huge rolling rack, so probably +/- at least several Hz, which is considered inaudible. Also, I was measuring dual 20 Hz Fs drivers = Fb tuning in ~30 ft^3 referenced to 40 Hz, about the lowest signal on vinyl at the time, so ~ -24 dB.
In short, way too out of date to be of any use nowadays.
GM
Indeed! Bjorno was quite the complete pro designer with the necessary math skills to work it out using the wave equation and the tools/personnel at Dynaudio? to go way beyond my mostly Altec 'enhanced' DIY efforts and somewhat of an advanced mentor for me, but sadly 'left the building' recently.
1] Correctomundo!
2, 3, 4] Not anytime soon, if ever. Ages ago I figured them all out based on Altec data [actually Western Electric's] and at this late date just remember they didn't agree WRT any previously slot/shelf port design theory, but it's all on two damaged HDs, so probably lost to me and without the data can't compare to this one or try again as it was absorbed by EV and they don't share, at least to me.
I did notice though that he changed/added a significant amount of info from earlier versions of his and others and had planned to discuss it with him after the holidays, especially #4 since Altec lumped them together as one, though used thin mica dividers to mimic WE's metal ones with a high aspect ratio end correction [added friction, re heat/AC duct design], though wondering if NACA duct design might be a better choice, so with typical ~3/4" thick defining square to modest rectangular [1:1.273 Altec max] 1/2 WL open pipes it seems reasonable that it might be more like his .......... regardless, based on who made it, I'm inclined to accept this latest one as at least 'close enough' till proven otherwise.
With mica dividers, no measurable difference Vs no dividers with the same tuning or round vent of the same area, though need to keep in mind that my measurement capability was only a notch above none compared to the Altec distributor's $20+ k [early '70s dollars] huge rolling rack, so probably +/- at least several Hz, which is considered inaudible. Also, I was measuring dual 20 Hz Fs drivers = Fb tuning in ~30 ft^3 referenced to 40 Hz, about the lowest signal on vinyl at the time, so ~ -24 dB.
In short, way too out of date to be of any use nowadays.
GM
Thanks again!
Ever consider sending your HDs out for data recovery? I've been lucky over the decades and only suffered a catastrophic loss on the single 5-1/4" floppy disk I'd stored all of my code on when I was 14. But I know there are companies that offer those services.
I get the impression that the standard port equation and end corrections are still missing something. I saw a fairly good experiment (vent tuning) looking at free vs baffle vs flare+baffle with round ports that still had length errors of 30-40% between theory and measurement. The k values were in the noise.
This article (End correction at flue pipe mouth) on flue pipe mouth length tuning was interesting, and these look like the inside end of a simple slot port (to me).
Finally, the fact that I've been digging deep into this for only a month or two while the authors of some of the software out there have been at it for a couple orders of magnitude longer than that makes me want to use the software. That's why I'm on the hunt for a Boxsim tutorial. AKABAK looks awesome, too.
Sorry to hear about your mentor.🙁
Ever consider sending your HDs out for data recovery? I've been lucky over the decades and only suffered a catastrophic loss on the single 5-1/4" floppy disk I'd stored all of my code on when I was 14. But I know there are companies that offer those services.
I get the impression that the standard port equation and end corrections are still missing something. I saw a fairly good experiment (vent tuning) looking at free vs baffle vs flare+baffle with round ports that still had length errors of 30-40% between theory and measurement. The k values were in the noise.
This article (End correction at flue pipe mouth) on flue pipe mouth length tuning was interesting, and these look like the inside end of a simple slot port (to me).
Finally, the fact that I've been digging deep into this for only a month or two while the authors of some of the software out there have been at it for a couple orders of magnitude longer than that makes me want to use the software. That's why I'm on the hunt for a Boxsim tutorial. AKABAK looks awesome, too.
Sorry to hear about your mentor.🙁
You're welcome!
Yes, was quoted $120/hr in 2010? and wasn't hopeful since a motherboard short caused a fire while doing a BU, hence two damaged HDs. A local was able to retrieve the 'D'? partition of the BU, but most of the good stuff [all the many/varied programs + work product was on 'C', which couldn't be accessed. The main HD wouldn't do anything.
A few folks around the net offered to take a look-see at the time, but none local and frankly, at this late date and other reasons have for the most part lost interest beyond helping me to remember stuff I use to take for granted.
GM
Yes, was quoted $120/hr in 2010? and wasn't hopeful since a motherboard short caused a fire while doing a BU, hence two damaged HDs. A local was able to retrieve the 'D'? partition of the BU, but most of the good stuff [all the many/varied programs + work product was on 'C', which couldn't be accessed. The main HD wouldn't do anything.
A few folks around the net offered to take a look-see at the time, but none local and frankly, at this late date and other reasons have for the most part lost interest beyond helping me to remember stuff I use to take for granted.
GM
